© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129435
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on tokens that kick off or emphasize a personal or subjective opinion—especially first‐person pronouns (“I,” “It” when used subjectively) or related evaluative remarks.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    sentence ending punctuation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    <>
    -1.09
     jaj
    -0.96
     atmosfer
    -0.92
    ❙
    -0.90
     neceſſ
    -0.90
    😭
    -0.89
    😞
    -0.88
     جدید
    -0.88
     baru
    -0.85
     تازه
    -0.85
    POSITIVE LOGITS
     overall
    1.22
    Overall
    0.99
    Had
    0.94
    Zubereitung
    0.93
    えたら
    0.90
     would
    0.89
    Still
    0.88
     I
    0.88
    矛盾
    0.88
     will
    0.87
    Activations Density 0.007%

    No Known Activations